Botulinum neurotoxins (BoNTs) are important medical and cosmetic agents, used to treat dystonias, blepharospasms, hyperhidrosis, and other neurological diseases. However, BoNTs also represent the most toxic substances known and their potential abuse as a threat agent is feared (Arnon 2001; Wein 2005). The detection of Botulinum neurotoxin (BoNT) in complex samples such as foods or clinical specimens represents an analytical challenge. The current “gold standard” in the art for detecting BoNT is the mouse toxicity assay, which can detect as little as 10 pg BoNT (Ferreira 2003). However, BoNT can be lethal to humans in systemic doses as low as 1 to 2 ng/Kg body weight (Arnon 2001). Therefore, there is a need in the art for more sensitive assays for detecting the presence of BoNT in a sample.
BoNTs have gained popularity as cosmetic drugs, and have also been successfully used for the treatment of a variety of neurological and neuromuscular disorders (Schantz 1992; Johnson 1999). Products containing BoNT are approved for the medical treatment of several diseases, i.e. cervical dystonia, torticolis, blepharospasm, hyperhydrosis, strabismus and migraines. Further, the products BOTOX COSMETICS®, VISTABEL®, BOCOUTURE® and AZZALURE® are licensed as an anti-wrinkle treatment. With the ever-increasing medical use of BoNT, its sensitive and specific detection in manufacturing processes as well as clinical research laboratories is of crucial importance. Accordingly, the potency of each batch of BoNT must be determined by the manufacturer before release to ensure the safety and efficacy of the product. This testing is widely performed using the classical mouse LD50 assay, which measures the potency of each batch of the product by determining the dose that will kill 50 percent of the animals. Animal testing for each batch of BoNT is expensive, slow, provides limited throughput, and requires sacrificing animals. Also, because of the lack of a standardized testing procedure, the units of biological activity are often unable to be directly converted into precise doses for human use, and overtreatment with BoNTs can cause iatrogenic forms of botulism (Partikian 2007; Crowner 2007). Thus, there is a need for an alternative method for testing products containing BoNT that is more efficient and cost effective.
Natural BoNT resides within ˜300, 500 or 900-kDa protein complexes together with other non-toxic components, the neurotoxin associated proteins (NAPs) (Sakaguchi 1982; Chen 1998; Sharma 2003; Melling 1988; Zhang 2003; Aoki 2001). Several structurally distinct serotypes of BoNT (types A to G) have been discovered. BoNT Type A (BoNT/A) is most prevalent in the Western United States (Smith 1978) and is causatively involved in approximately 60% of the IB cases in California (the rest being mostly attributed to type B) (Arnon 2001). The toxin itself is a 150-kDa zinc-binding metalloprotease that, following expression, is endogenously cleaved into a 100-kDa heavy and a 50-kDa light chain connected by a reducible disulphide bond (Schiavo 2000) and by a belt-like extension of the heavy chain that loops around the light chain (Lacy 1998). The catalytic site is located on the light chain (Kurazono 1992). Reduction of the chain-bridging disulphide bond exposes the catalytic site and enhances its activity (Lacy 1998), also referred to as “activation” of the toxin by some authors and toxin manufacturers (Cai 1999; Cai 2001). The potency of BoNT results from its ability to cleave on or more of the three SNARE proteins involved in fusing acetylcholine-containing synaptic vesicles with terminal motor neurons membrane, triggering muscle contraction (Shiavo 2000).
Detection of low levels of BoNT in a sample using prior art methods is difficult. However, due to the enormous potency of the toxin, which can be lethal for humans in systemic doses of 1 to 2 ng/Kg body weight (Arnon 2001), these low levels can be extremely dangerous. For example, in infant botulism (IB), a condition in which a baby's intestines have become colonized by toxin-secreting Clostridium botulinum bacteria, it is possible to detect BoNT in stool samples (Arnon 2006). However, attempts to diagnose IB serologically via detection of BoNT in the blood have been deemed unreliable (Schantz 1992). Nevertheless, the systemic presence of the toxin in IB cannot be disputed, because of its apparent quick distribution throughout the infant's entire body, by which it efficiently shuts down motor neurons distant from the intestinal source. The resulting symptoms can include complete paralysis and respiratory failure.
The definite diagnosis of botulism requires detection of BoNTs in clinical specimens. Most commonly used and relied on is the life mouse assay. This assay can detect as little as 10 pg BoNT (Ferreira 2003). In the life mouse assay, mice are injected intraperitoneally (i.p.) with 0.5 mL/mouse of sample, treated with type A or B antitoxin, and observed for signs of botulism or death, typically over a 48 hour period. Toxicity is expressed by the number of hours until death (Kautter 1977; Sharma 2006). As in many animal experiments, the results of the mouse assay may vary. Four- to five-fold differences in response to a given dose are typical (Sugiyama 1980). Other and generally faster methods for BoNT detection include use of fluorescence resonance energy transfer (FRET) substrates for BoNT (U.S. Pat. No. 6,504,006), various enzyme-linked immunosorbent assays (ELISAs) (Sharma 2006), Enzyme-amplified protein micro arrays with a “fluidic renewable surface fluorescence immunoassay” (Varnum 2006), mass spectrometric assays (Barr 2005; Kalb 2005; Boyer 2005; Kalb 2006), immuno-PCR detection (Chao 2004), and recently, a real-time PCR-based assay that utilizes reporter DNA-filled liposomes which bind to immobilized BoNT/A via gangliosides (Mason 2006a; Mason 2006b). Reported detection limits and sample types for these various methods are summarized in Table 1. Except for the PCR-based assays, most assays are not well suited to provide the desired detection of less than 1 pg/mL BoNT in a complex sample. By approximation, 1 pg/mL corresponds to the lethal concentration under presumed equal distribution throughout the human body.
TABLE 1Reported performance of existing Botulinum toxin assaysDemonstrated forSensitivity AssayTest methodSample Type(fg/mL)TimeMass spectrometrymilk, serum, stool320,000   <4 hrs(Endopep-MS)22-25extractEnzyme-linkedliquid and solid60,000  6-8 hrsimmunosorbent assaysfoods, serum(ELISA)19ELISA-HRP29therapeutic9,000  4-6hrspreparationsMouse assay (goldfoods, serum, stool ~6,000    typically 48standard)18hrsEnzyme-amplified proteinblood, plasma1,400  <10 min. permicroarray and fluidicmeasurementrenewable surfacefluorescenceimmunoassay21Immuno-PCR26carbonate buffer 504-6 hrsImmuno-PCR withdeionized water    0.026 hrsganglioside-mediatedliposome capture27, 28
Anthrax lethal factor (LF) is another zinc metalloprotease that has been successfully been adapted for use in the ALISSA. LF constitutes one of the three components of anthrax toxin that is produced by Bacillus anthracis, together with protective antigen (PA) and edema factor (EF) (Brossier 2001). LF specifically cleaves members of the mitogen-activated protein kinase kinase (MAPKK) family, leading to the inhibition of essential signaling pathways. LF alone is not toxic; it requires the presence of PA for its translocation into cells (Brossier 2001). Macrophages are believed to be primarily affected by LF (Hanna 1993). A specific and sensitive assay for the detection of LF is potentially useful for early diagnosis of anthrax infection and is expected to be a useful research tool to advance the understanding of the mechanism of action of anthrax toxin (Boyer 2007).